Improving Simulation-Based Origin-Destination Demand Calibration Using Sample Segment Counts Data

TL;DR

The paper addresses underdetermination in origin-destination (OD) demand calibration for high-resolution stochastic traffic simulators by introducing a regularization term based on segment-level sample counts. It extends the metamodel-based simulation-based optimization framework to incorporate segment-count regularization while preserving a path-travel-time misfit objective, yielding a robust dual-objective formulation. In a Seattle highway network study, the regularized approach substantially improves recovery of ground-truth OD and segment demand, achieving up to a 65% reduction in the normalized RMSE with only modest travel-time impact. The method is scalable and adaptable to diverse trajectory data sources, offering a principled approach to leverage high-resolution surveillance for improved demand calibration.

Abstract

This paper introduces a novel approach to demand estimation that utilizes partial observations of segment-level track counts. Building on established simulation-based demand estimation methods, we present a modified formulation that integrates sample track counts as a regularization term. This approach effectively addresses the underdetermination challenge in demand estimation, moving beyond the conventional reliance on a prior OD matrix. The proposed formulation aims to preserve the distribution of the observed track counts while optimizing the demand to align with observed path-level travel times. We tested this approach on Seattle's highway network with various congestion levels. Our findings reveal significant enhancements in the solution quality, particularly in accurately recovering ground truth demand patterns at both the OD and segment levels.

Figures (1)

• Figure 1: Comparison of the baseline and proposed regularized solver performance across different metrics and scenarios